I agree 100% with Pekka's assessment of the current state of interoperability testing, reporting and requirements as applied to advancement of protocol standards.
I also agree that we should be more precise in our acceptance of "implemented and interoperable". I fear that the current practice can be (and has been) applied selectively to allow advancement of some standards while holding back others.
- Ralph
At 02:32 PM 4/27/2004 +0300, Pekka Savola wrote:
On Tue, 27 Apr 2004, JINMEI Tatuya / [ISO-2022-JP] [EMAIL PROTECTED]@C#:H(B wrote: > In any event, I'd first like to clarify the general point before going > to each detail to avoid further confusion. The question is: > > We do not have an implementation on some part of RFC2462. Can we > still recycle rfc2462bis as DS (process-wise), keeping that part, > despite the lack of implementation? > > If the answer is yes, we can concentrate on technical aspects of the > discussion for both the M and O flags. > > If the answer is no, we need to somehow deprecate/remove the M flag, > and then concentrate on issues about the O flag (as you suggested > above).
This is a good question, and one for which you're not likely to get a definitive answers.
In the past the implementation & interoperability reports have been more or less half-assed, and you could not really get a good feel on what's actually implemented and tested to interoperate. I've personally sensed a desire for a bit more detail, but not necessarily to all the gritty details.
On the other hand, in Robert Elz's appeal, at least in that case, the IESG felt (or that's how I interpreted it at least) that it is traditionally not been required to demonstrate implementation and interoperability of even the smallest details of the spec:
http://www.ietf.org/IESG/APPEALS/kre-ipngwg-addr-appeal-withdraw.txt
[...] The IESG notes that the existing practice when advancing documents on the standards track has not involved having implementation reports include detailed verification that implementations enforce every statement as is implied above, in the absence of explicit text requiring that an implementation make such checks. [...]
So, my personal feeling is, "yes, we should be stricter on what we accept as "implemented & interoperable", but whether that happens in practice is another thing.
Hope this helps..
-- Pekka Savola "You each name yourselves king, yet the Netcore Oy kingdom bleeds." Systems. Networks. Security. -- George R.R. Martin: A Clash of Kings
-------------------------------------------------------------------- IETF IPv6 working group mailing list [EMAIL PROTECTED] Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------
-------------------------------------------------------------------- IETF IPv6 working group mailing list [EMAIL PROTECTED] Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6 --------------------------------------------------------------------
